|
Brian Keller
Hi. My name is Brian Keller. I am a PhD Candidate in Systems
and Industrial Engineering at the University of Arizona. I am interested
in operations research modeling and optimization. You can contact
me at kellerbd@email.arizona.edu.
|
|
PhD, Systems and Industrial Engineering, University of Arizona, Tucson, AZ (Spring 2009)
(Minor in Finance)MS, Systems Engineering, University of Arizona, Tucson, AZ (2005)
BS, Physics and Business, Mount Union College, Alliance, OH (2004)
Research Interests
I am interested in operations research modeling with stochastic components. Financial modeling, revenue management, and scheduling are some areas I am exploring. My dissertation research is in stochastic integer programming with applications in team scheduling and production scheduling. I am working with Dr. Guzin Bayraksan.
Projects
Generating Flat Radar Waves
Optimizing component placement in a radar system can generate flat waves close to the radar. Flat waves are characterized by flat amplitude and phase. Since the problem was highly complex and unstructured, we developed a randomized greedy search to find component placements that would generate the flat waves. This work was done for the Waveband Corporation.
Tabu Search
Want to learn more about Tabu Search? Click here to see my Tabu Search links.
Independent Consultant for Code Denver - 2007
I developed an integer program and team pairing hueristic to schedule soccer leagues. There were some interesting constraints such as some coaches may coach more than one team and these coaches would like to play their games back-to-back. The model is being used to schedule leagues of up to ten divisions with up to 26 teams per division and has saved the client $30,000 per year.
Key Bank, Cleveland, OH - Summer 2006
I worked in the Quantitative Risk Analysis group at Key Bank focusing on operational risk. Key Bank's assets total more than $100 billion. Operational risk is the risk of loss due to internal fraud, external fraud, workmen's compensation, litigation, etc. I debugged and validated the bank's proprietary C++ math library used to calculate economic capital. Economic capital is the amount of money the bank must set aside to cover unexpected losses. I also conducted a study to determine under what conditions a closed-form approximation accurately calculated the economic capital.
Research Assistant
- Working with Dr. Guzin Bayraksan researching stochastic integer programs with applications in team scheduling and production scheduling.
- Worked with Dr. Leo Lopes on the Flat Wave Project
Teaching Assistant
- SIE 530 Engineering Statistics
- SIE 500a, 500b, 500c Introduction to SIE Methods, course instructor
- ENGR 256 Engineering Economic Analysis
- ENGR 170 Programming in C